| Type: | Package |
| Version: | 0.2 |
| Title: | Threshold Cut Point of Probability for a Binary Classifier Model |
| Date: | 2017-09-02 |
| Author: | Navinkumar Nedunchezhian |
| Maintainer: | Navinkumar Nedunchezhian <navinkumar.nedunchezhian@gmail.com> |
| Description: | Allows to view the optimal probability cut-off point at which the Sensitivity and Specificity meets and its a best way to minimize both Type-1 and Type-2 error for a binary Classifier in determining the Probability threshold. |
| License: | GPL-2 |
| LazyData: | FALSE |
| Imports: | ggplot2,reshape2 |
| Suggests: | knitr |
| NeedsCompilation: | no |
| Packaged: | 2017-09-02 16:27:17 UTC; NSD |
| Repository: | CRAN |
| Date/Publication: | 2017-09-02 17:27:38 UTC |
This Supports the datascientist to determine the optimal threshold for binary classifier problem by visuallizing the sensitivity, specificity and accurarcy of the given model
Description
Prints 'Chart of sensitivity & specificity'.
Usage
Binary_threshold(probability,class)
Arguments
probability |
Probability Obtained from the model |
class |
Actual Class of the datasets |
Examples
set.seed(100);disease <- sample(c("yes","no"), 1000, replace=TRUE);
Probabilities<-sample(seq(0,1,by=0.01),1000,replace=TRUE);
Binary_threshold(Probabilities,disease)